Do Probing and Steering Favor the Same Directions Within and Across Language Models?
Abstract
The best direction for probing a concept in a language model's hidden state is not the best direction for steering the model toward that concept, with implications for transfer and abstention.
A direction in a language model's hidden state can be used to probe whether a concept is present or to steer the model toward it. Although probe directions are often reused for steering, the direction that best detects a concept need not be the one that best changes it. We study this distinction in three instruction-tuned models from different families. We construct a one-parameter family of directions ranging from a covariance-adjusted discriminant to the mean difference, and evaluate probing and additive steering, which adds the direction to the hidden state, along the family at a fixed side effect. For truthfulness and sentiment, in every model, the best probe is not the best steering direction: steering consistently favors directions closer to the mean difference. This separation holds across 59 of 60 layer-model-concept combinations in a layer sweep, with matched probe and steering texts in all six model-concept combinations, and under a rotation-based steering operator. The exact steering optimum is less stable: for sentiment, an intermediate direction produces a 27-56% stronger steering effect than the mean difference, but this gain shrinks or disappears with matched texts and depends on the layer and the operator. The two roles also transfer differently across models. A probe transferred by a linear map fitted on neutral text retains 81-91% of the target model's above-chance probing performance, while the transferred mean difference retains less of its steering effect in 11 of 12 model pairs. This is a consistent tendency, although our predefined criteria classify only three pairs, and nine remain inconclusive. Finally, as an application on PopQA, steering toward "unknown" increases abstention by 18-44 percentage points on uncertain errors, versus 2-5 on confident errors, while rarely correcting errors in cases where a correct answer is reachable.
Community
The direction that best detects a concept in an LLM is rarely the one that best steers it: 59 of 60 settings across 3 model families. Interactive explainer: https://www.afshn.com/research/probing-vs-steering/
What I like here is that probing and steering are treated as two different jobs, since reusing a probe direction to steer is such a common shortcut. Seeing the best detecting direction differ from the best steering one across three instruction-tuned models is a useful warning. Does the gap between them change with model size, or is it about the same in all three models?
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Interpretability Is Not Model-Specific (2026)
- Probing and Steering for Safety-Relevant Behaviours in Large Language Models (2026)
- One Vector or Many? Clustered Optimal-Transport Steering under Quantization (2026)
- When a Conceptor Complement Becomes Gain: Operator Geometry and Behavioral Attribution in Language Models (2026)
- Written Once, Read Differently How Language Models Own Answers Steer Their Next Ones (2026)
- A helps B while B hurts A: directed transfer in instruction-tuning mixture (2026)
- A Capitulation Direction Emerges With Scale and Is Causally Load-Bearing Where Validation Is Strong: Eight Models, Two Families, and Five Measurement Hazards (2026)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
Get this paper in your agent:
hf papers read 2611.19499 Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper